Fault diagnosis method for three-phase permanent magnet synchronous motor drive system based on model predictive control

Abstract

The application discloses a three-phase permanent magnet synchronous motor drive system fault diagnosis method based on model prediction control and relates to the field of permanent magnet synchronous motor fault diagnosis. The three-phase common-mode voltage is observed, the time point t1 of the occurrence of the switch tube fault is determined according to the three-phase common-mode voltage relationship, the waveform of the value function is observed, and the time point t2 of the first appearance of the wave crest of the value function is obtained; according to the phase difference between the time point t1 and the time point t2, the fault switch tube is positioned through the comparison table of the phase difference and the fault switch tube. The amplitude and the phase of the value function waveform are used for judgment, so that the fault diagnosis and positioning of the switch tube are realized.

Description

Technical Field

[0001] This invention relates to the field of fault diagnosis of permanent magnet synchronous motors, and in particular to a fault diagnosis method for a three-phase permanent magnet synchronous motor drive system based on model predictive control. Background Technology

[0002] With societal development, permanent magnet synchronous motors (PMSMs) have been widely used in numerous fields, facing increasingly complex operating environments. For example, they often require prolonged operation in high-temperature and humid environments, which frequently leads to PMSM drive system failures. Failure to detect these failures promptly can cause irreversible damage to the motor and even endanger human life and property. Therefore, researching fault diagnosis strategies for three-phase permanent magnet synchronous motors (PMSMs) has significant engineering importance and value. Among the various fault types in PMSMs, mechanical and electrical faults are the most common. Among electrical faults, open-circuit faults in the drive system's switching transistors are particularly prevalent, and fault diagnosis of open-circuit faults in the drive system's switching transistors should be considered. Currently, a dual closed-loop control strategy is commonly used in three-phase motor drive systems. In comparison, MPCC control is simpler, as the current, the control variable, can be directly measured through a sampling circuit, and the constructed value function is only related to the stator current under the dq axis. When using MPCC control, an open-circuit fault in the drive system's switching transistors causes a linkage effect; that is, an open circuit in one transistor affects the switching signals of other transistors, thus impacting the stability of the entire system. In MPCC-controlled three-phase motor drive systems, existing fault diagnosis methods primarily utilize two key pieces of information: switching quantities and the cost function. Faults are detected by monitoring the amplitudes of the DC and second harmonic components in the cost function, or by further studying the cost function using Fourier series methods. These diagnostic methods obtain specific values ​​from the cost function to detect faults, and then combine them with other methods to locate the fault. However, the cost function is highly correlated with motor parameters and sampling time. The accuracy of the diagnosis depends on precise parameters and appropriate predefined thresholds. Changes or mismatches in parameters will alter the output of the cost function, leading to misdiagnosis. Therefore, researching how to quickly and accurately locate faulty switching transistors when a fault occurs in a three-phase permanent magnet synchronous motor drive system has significant application significance and practical value. Summary of the Invention

[0003] This invention provides a fault diagnosis method for a three-phase permanent magnet synchronous motor drive system based on model predictive control, comprising the following steps:

[0004] Observe the three-phase common-mode voltage and determine the time point t1 when the switching transistor fault occurs based on the relationship of the three-phase common-mode voltage.

[0005] Observe the waveform of the value function and obtain the time point t2 when the value function first appears at its peak;

[0006] The phase difference is calculated based on time points t1 and t2, and the faulty switch is located by referring to the phase difference and the faulty switch.

[0007] Furthermore, the three-phase common-mode voltage is:

[0008] ;

[0009] in, , and These are the common-mode voltages of phases a, b, and c, respectively. , and This is the reference value for the phase voltage output of the inverter; , and Leakage inductance voltage drop; , and Represents three back electromotive forces; when , and Not satisfied = = The time at which the switching transistor failure occurs is the time at which the failure occurs.

[0010] Furthermore, the reference voltage , and They are respectively:

[0011] ;

[0012] Among them, S a S b S c These represent the switching states of phase a, phase b, and phase c bridge arms, respectively.

[0013] Furthermore, the method for obtaining the leakage inductance voltage drop includes:

[0014] Establish a differential tracker:

[0015] ;

[0016] in, The input signal to be differentiated is a three-phase current. , and , To obtain the differential result, we get , and ;

[0017] Through respectively The leakage inductance voltage drop of the motor can then be obtained.

[0018] Furthermore, the method for obtaining the back electromotive force includes:

[0019] Two extended state observers are set up to observe the back potential. , ,for:

[0020] ;

[0021] Where k is the gain coefficient. ; , , and sgn() represents the differential value of the αβ axis current;

[0022] Three-phase back potentials are obtained using the inverse Clark transform. , and for:

[0023] .

[0024] Furthermore, the differential value of the αβ axis current and for:

[0025] ;

[0026] in: , These are the d-axis and q-axis inductance components, respectively. For mechanical angular velocity, The mechanical angle through which the rotor rotates. It is a permanent magnet flux linkage. , The d-axis and q-axis components of the stator flux linkage are:

[0027] ;

[0028] ;

[0029] , .

[0030] Furthermore, the value function is used to represent the following relationship between the current of the motor stator and the set reference current value at the current moment. The value function is:

[0031] ;

[0032] in, This represents the reference current along the d-axis in a synchronously rotating coordinate system. This represents the reference current along the q-axis in a synchronously rotating coordinate system. and This represents the current along the d-axis and q-axis for the (k+1)th time.

[0033] Furthermore, the (k+1)th current along the d-axis and q-axis and for:

[0034] ;

[0035] in, and Let be the d-axis and q-axis currents for the k-th current.

[0036] Compared with the prior art, the present invention has the following technical effects:

[0037] 1. This invention is based on the open-circuit fault of the switching transistor in a three-phase permanent magnet synchronous motor drive system. Using a simplified model of the three-phase permanent magnet synchronous motor drive system, the common-mode voltage equation of the system is re-derived. By utilizing the characteristics of the common-mode voltage changes under normal and fault conditions, an open-circuit fault can be detected within one switching cycle, giving the system a faster fault detection capability.

[0038] 2. This invention re-derives the relationship between the switching combination and voltage after a fault. It reconstructs the current prediction control function after a fault and locates the faulty switching transistor by judging the amplitude and phase relationship of the MPCC output value function after the fault.

[0039] 3. This invention is applicable to applications with high system reliability requirements, such as electric vehicles, rail transit, and wind power generation, and has greater application value in high-load and high-speed applications.

[0040] The following will further explain the concept, specific structure, and technical effects of the present invention in conjunction with the accompanying drawings, so as to fully understand the purpose, features, and effects of the present invention. Attached Figure Description

[0041] Figure 1 This is a schematic diagram of the topology of the three-phase permanent magnet synchronous motor drive system of the present invention;

[0042] Figure 2 This is a schematic diagram of the predictive current control block of the model of this invention;

[0043] Figure 3 This is a topology diagram of the three-phase permanent magnet synchronous motor drive system when S1 is disconnected according to the present invention;

[0044] Figure 4 This is a waveform diagram of the value function output of the model predictive current control system when the system of the present invention is running normally;

[0045] Figure 5 This is the three-phase current waveform of the motor windings when the system of this invention is operating normally;

[0046] Figure 6 This is a waveform diagram of the value function output of the model predictive current control system when the inverter S1 has an open-circuit fault according to the present invention.

[0047] Figure 7 This is a waveform diagram of the three-phase current of the motor winding when the drive system S1 of this invention fails;

[0048] Figure 8 This is a waveform diagram of the value function output of the model predictive current control system when the inverter S2 has an open circuit fault according to the present invention;

[0049] Figure 9 This is a waveform diagram of the three-phase current of the motor winding when the drive system S2 of this invention fails. Detailed Implementation

[0050] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0051] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the illustrations only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0052] Some exemplary embodiments of the invention have been described for illustrative purposes. It should be understood that the invention may be implemented in other ways not specifically shown in the accompanying drawings.

[0053] like Figure 1 As shown, a fault diagnosis method for a three-phase permanent magnet synchronous motor drive system based on model predictive control includes the following steps:

[0054] S1. Observe the three-phase common-mode voltage and determine the time point t1 when the switching transistor fault occurs based on the relationship of the three-phase common-mode voltage.

[0055] S11. Obtain the common-mode voltage mathematical model based on the drive system of the three-phase permanent magnet synchronous motor;

[0056] In one specific embodiment, the following is adopted: Figure 1 The schematic diagram of the three-phase permanent magnet synchronous motor drive system topology shown is illustrated. Based on the equivalent simplified circuit, the three common-mode voltages can be obtained from the switching state and phase current of each switching cycle:

[0057] (1)

[0058] in, This is the reference value for the inverter output phase voltage. When the system is operating normally, the inverter output phase voltage satisfies the following relationship:

[0059] (2)

[0060] S12, Definition The switching states of the three-phase bridge arms are used to obtain the reference voltage. ;

[0061] The switching states of the three-phase bridge arms are as follows:

[0062] (3)

[0063] Reference value of inverter output phase voltage This is directly related to the switching state in each cycle under the MPCC strategy, and the reference voltage. for:

[0064] (4)

[0065] S13, Based on the motor stator resistance R s and three-phase current , and Determine the stator resistance voltage drop of the motor as follows: , and ;

[0066] S14. Calculate the leakage inductance voltage drop of the motor based on the phase current differential information;

[0067] In one specific embodiment, to reduce the impact of measurement noise on differential information, a nonlinear differential tracker is used to obtain the differential information of the current. The basic equation of the differential tracker is as follows:

[0068] (5)

[0069] in, The input signal to be differentiated is a three-phase current. To obtain the differential result, It is in equation (9) , , Obtain the differential information of the current, respectively through The leakage inductance voltage drop of the motor can then be obtained.

[0070] S15. Use an extended state observer to observe back EMF information;

[0071] In this embodiment, a permanent magnet synchronous motor is established in... The state equation in the coordinate system is:

[0072] (6)

[0073] in: ;

[0074] ;

[0075] , These are the d-axis and q-axis inductance components, respectively. For mechanical angular velocity, , These are the d-axis and q-axis components of the stator flux linkage. The mechanical angle through which the rotor rotates. It is a permanent magnet flux linkage. , .

[0076] Two extended state observers are set up to observe the back potential. , ,for:

[0077] (7)

[0078] Where k is the gain coefficient. , , , , The value obtained by the sgn function is multiplied by the gain coefficient k; sgn() represents the sign function.

[0079] Three-phase back potentials are obtained using the inverse Clark transform. , and for:

[0080] (8)

[0081] S16. Based on the above reference voltage, resistor voltage drop, leakage inductance voltage drop, and back electromagnetism, the common-mode voltage is obtained as follows:

[0082] (9)

[0083] in, For leakage sensation, , , The three-phase current is obtained by differentiating the differential equation established by formula (5).

[0084] When an open-circuit fault occurs in the system, the above-mentioned common-mode voltage calculation formula no longer holds. For example... When an open circuit fault occurs, At that time, the actual voltage is equal to Therefore, it will be less than its reference voltage. The voltage relationship will become:

[0085] (10)

[0086] It is evident that the new common-mode voltage model can promptly reflect the fault conditions of the switching transistor. Based on the changes in the common-mode voltage, it can quickly identify whether an open-circuit fault of the switching transistor has occurred, and record the time point at which the open-circuit fault of the switching transistor occurs as t1.

[0087] S2. Observe the waveform of the value function and obtain the time point t2 when the value function first appears at its peak;

[0088] Establish a predictive control function for a three-phase permanent magnet synchronous motor drive system model;

[0089] In a synchronous rotating coordinate system, the stator voltage equation of the PMSM can be expressed as:

[0090] (11)

[0091] The model predictive current control selects the stator current of the permanent magnet synchronous motor as the controlled variable, establishes a predictive current model, and transforms the stator voltage equation of the three-phase permanent magnet synchronous motor in the dq axis of the synchronous coordinate system into the current differential equation in the dq axis, as follows:

[0092] (12)

[0093] Transforming it into matrix form is as follows:

[0094] (13)

[0095] According to the Euler equation of the prediction model, we can obtain:

[0096] (14)

[0097] Discretizing the current equation yields the following predictive current model:

[0098] (15)

[0099] in, and This represents the current at the k-th time. and This represents the (k+1)th current. Indicates the sampling time. This represents electric angular velocity. This represents the permanent magnet flux linkage in an electric motor.

[0100] In MPCC, a value function is needed to ensure that the measured motor stator current follows the set reference current value. The value function is defined as follows:

[0101] (16)

[0102] in, This represents the reference current along the d-axis in a synchronously rotating coordinate system. This represents the reference current along the q-axis in a synchronously rotating coordinate system. It is a nonlinear function used to limit the amplitude of the stator current, which is 0 or infinite within a certain range. i The subscript i indicates that this value function is the value function of the model predictive current control. According to equation (16), under normal circumstances, the difference between the actual value and the estimated value of the dq axis current is very small, and the value function is close to zero.

[0103] In a three-phase, two-level permanent magnet synchronous motor control system, there are a total of eight basic voltage state vectors, resulting in eight vectors in the value function. The last term in the equation is used to limit the stator current amplitude. The value function will undergo specific changes after an open-circuit fault occurs in the switching transistor. Taking an open-circuit fault in transistor T1 as an example, substituting the eight basic voltage vectors into Table 1 yields the voltage relationships under fault conditions.

[0104] Table 1:

[0105]

[0106] When the upper arm of phase A has no signal and is not working, but the lower arm has a signal (i.e., under normal circumstances), the AC phase voltage is measured. The relationship with the switching function is as follows:

[0107] (17)

[0108] After the Clark transformation, we get:

[0109] (18)

[0110] Combining the two equations above, we obtain the switch state and... The relationship is as follows:

[0111] (19)

[0112] Applying the Park transformation to the above equation yields:

[0113] (20)

[0114] Right now:

[0115] (twenty one)

[0116] When the upper bridge arm switch of phase A should be in the on / off state and the lower bridge arm off, and when the upper bridge arm switch of phase A is in the open circuit state, then the upper and lower bridge arms of phase A are essentially completely disconnected. Under this fault condition, the AC phase voltage... Stator voltages along the α and β axes Stator voltage of the dq axis The relationship with the switching function is as follows:

[0117] (twenty two)

[0118] After Clark transform, we can obtain:

[0119] (twenty three)

[0120] Combining the two equations above, we get:

[0121] (twenty four)

[0122] Performing the Park transformation on the above equation yields:

[0123] (25)

[0124] Right now:

[0125] (26)

[0126] After obtaining the voltage equation after an open-circuit fault, substitute it into the current prediction model in formula (15) to obtain the reconstructed current prediction model, and then obtain the cost function under the fault condition.

[0127] This invention reconstructs the model predictive current control function after a fault. Analysis of the value functions of the model predictive current control before and after the fault reveals that when an open-circuit fault occurs, the actual switching signal is inconsistent with the controller's switching signal. This is because an incorrect command signal is used to estimate i. d (k+1) and i q (k+1), which is related to the given d-axis current. and q-axis current setpoint Excessive difference causes the value function to no longer be zero, changing it from a state where the value function is always zero to a jumping signal. Simulation results show that when different switching transistors have open-circuit faults, their value functions exhibit different phases of periodic changes. By judging the amplitude and phase of the value function, the fault location of the switching transistor can be achieved.

[0128] S3. Locate the faulty switch tube based on the phase difference between time point t1 and time point t2.

[0129] Common-mode voltage information exhibits different characteristics under normal and fault conditions. Derivation of the mathematical models for these two conditions reveals that the stator voltage under fault conditions differs. Under normal operation, the common-mode voltage of the three-phase system is identical, with a constant amplitude. If an open-circuit switch occurs, the common-mode voltage of all three systems will change abruptly. This allows for monitoring of system faults and recording the fault occurrence time. Since the motor side remains unchanged during a fault, the mathematical model of the PMSM (Motor-Modulated Switching System) remains unchanged, meaning the MPCC (Multi-Phase Control Center) program remains the same. However, the phase voltage on the AC side of the motor differs depending on the faulty switch, causing the value function in the MPCC to exhibit phase and amplitude differences depending on the faulty switch. Based on the fault time point given by the common-mode voltage, the phase difference when different switches fail can be obtained, allowing for the establishment of a lookup table between phase differences and faulty switches. Using time points t1 and t2 obtained in steps S1 and S2, the phase difference of the value function waveform at these two time points is determined, and the faulty switch is located using the aforementioned lookup table.

[0130] To further illustrate the present invention, in one specific embodiment, a structure is built as follows: Figure 2 The fault diagnosis model shown uses a three-phase permanent magnet synchronous motor with motor parameters as shown in Table 2. The motor is started at a given speed of 500 r / min under no-load conditions. The S1 open-circuit fault is simulated by cutting off the Sa+ signal for verification.

[0131] Table 2

[0132]

[0133] Figure 4This is the waveform of the value function output by the MPCC when the inverter is operating normally. Figure 5 The value function of the model predictive current control algorithm output is always 0 when the motor is running normally, which is the three-phase stator winding current of the motor.

[0134] Figure 6 The waveform of the value function output by the model predictive current control algorithm is shown when the switching transistor S1 in the motor drive system topology experiences an open-circuit fault in 0.1s. Figure 7 When the motor drive system S1 is open-circuited, the current of the three-phase stator winding of the motor is shown to be significantly oscillating. The system is set to experience an open-circuit fault at 0.1s. It can be seen that the value function changes significantly after the fault occurs at 0.1s. The value function first peaks at 0.105s and exhibits obvious periodicity.

[0135] Figure 8 The waveform of the value function output by the model predictive current control algorithm when the switch S2 has an open-circuit fault is shown. Using the peak detection function built into MATLAB, the first peak time of the value function can be found to be 0.108s. The phase of the value function has similar characteristics when other switches have faults. Therefore, this feature of the value function can be used to diagnose and locate faulty switches.

[0136] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A fault diagnosis method for a three-phase permanent magnet synchronous motor drive system based on model predictive control, characterized in that, Includes the following steps: Observe the three-phase common-mode voltage, identify whether an open-circuit fault has occurred based on the change in common-mode voltage, and record the time point at which the open-circuit fault occurs as t1. The three-phase common-mode voltage is: ； in, , and These are the common-mode voltages of phases a, b, and c, respectively. , and This is the reference value for the phase voltage output of the inverter; , and Leakage inductance voltage drop; , and Represents three back electromotive forces; when , and Not satisfied = = The moment when the switching transistor failure occurs; Observe the waveform of the value function and obtain the time point t2 when the value function first appears at its peak. The value function is used to represent the following relationship between the current of the motor stator and the set reference current value. The value function is: ； in, This represents the reference current along the d-axis in a synchronously rotating coordinate system. This represents the reference current along the q-axis in a synchronously rotating coordinate system. and This represents the current along the d-axis and q-axis at the (k+1)th time. The phase difference is calculated based on time points t1 and t2, and the faulty switch is located by referring to the phase difference and the faulty switch.

2. The fault diagnosis method for a three-phase permanent magnet synchronous motor drive system based on model predictive control according to claim 1, characterized in that, The phase voltage reference value , and They are respectively: ； Among them, S a S b S c These represent the switching states of phase a, phase b, and phase c bridge arms, respectively.

3. The fault diagnosis method for a three-phase permanent magnet synchronous motor drive system based on model predictive control according to claim 1, characterized in that, The method for obtaining the leakage inductance voltage drop includes: Establish a differential tracker: ； in, The input signal to be differentiated is a three-phase current. , and , To obtain the differential result, we get , and ; Through respectively The leakage inductance voltage drop of the motor can then be obtained.

4. The fault diagnosis method for a three-phase permanent magnet synchronous motor drive system based on model predictive control according to claim 1, characterized in that, The method for obtaining the back electromotive force includes: Two extended state observers are set up to observe the back potential. , ,for: ； Where k is the gain coefficient. ; , , and sgn() represents the differential value of the αβ axis current; Three-phase back potentials are obtained using the inverse Clark transform. , and for: 。 5. The fault diagnosis method for a three-phase permanent magnet synchronous motor drive system based on model predictive control according to claim 4, characterized in that, Differential value of αβ axis current and for: ； in: , These are the d-axis and q-axis inductance components, respectively. For mechanical angular velocity, The mechanical angle through which the rotor rotates. It is a permanent magnet flux chain. , The d-axis and q-axis components of the stator flux linkage are: ； ； ， 。 6. The fault diagnosis method for a three-phase permanent magnet synchronous motor drive system based on model predictive control according to claim 1, characterized in that, The (k+1)th current along the d-axis and q-axis and for: ； in, and Let be the d-axis and q-axis currents for the k-th current.

Images